Getting GLContext on a Tesla

I have a Nvidia Tesla M2070Q installed on an IBM Blade server. The server is set with a Matrox video card as the primary video card. The Nvidia card does not have any external output.

Is it possible to get a OpenGL context from the Nvidia GPU? The OS that I used is Windows Server 2008 HPC Edition. I am primary trying to perform off-screen rendering (by reading off the p-buffer), thus a display is not necessary.

Thanks for helping.

Sounds like WGL_NV_gpu_affinity might be what you are looking for.

Have you thought about using FBOs instead of the old p-buffer stuff for off-screen rendering?

But I am unable to get this working, is there any help or sample code for this? It appears WGL_NV_gpu_affinity function is missing even after I installed the Nvidia OpenGL SDK and NVIDIA drivers.